Project Summary Tuberculosis is the leading cause of death by an infectious disease worldwide. Tuberculosis control strategies in low- and middle-income countries (LMICs), which bear the major burden of TB, currently suffer from two major gaps in the scientific evidence on program implementation, which must be overcome to achieve global targets: (1) how to trace the chain of transmission to the major pockets of the population where TB is increasingly concentrated (i.e., populations serving as reservoirs for broader population epidemics); and (2) how to efficiently and cost-effectively screen populations within these concentrated reservoirs. Our preliminary research and those of several other groups has revealed that prisons are likely to be an important reservoir for tuberculosis in many LMICs. Mass screening, as suggested by the World Health Organization, could be an effective means of case detection, but is not widely implemented in LMICs due to high costs and infrastructure requirements. We propose to leverage unique research and tuberculosis surveillance infrastructure in prisons and community settings in Central- Western Brazil to address these critical questions. We will test the following three hypotheses: 1) a major burden of tuberculosis in communities is attributable to transmission in prisons; 2) testing pooled sputum samples using a new, sensitive molecular diagnostic assay (Xpert Ultra) on a mobile diagnostic unit can accurately and efficient detect tuberculosis cases; and 3) prison-based mass screening can cost- effectively reduce the community burden of tuberculosis. This project will utilize a novel statistical modeling approach to infer the directionality of transmission by integrating tuberculosis natural history, exposure and phylogenetic data derived from whole genome sequencing. Additionally, this work will generate an open source tool for comparing the impact of various diagnostic strategies in prisons and other high-burden populations. Overall, this project addresses persistent scientific barriers to tuberculosis control: how to clearly identify the contribution of reservoir populations, and how to screen them efficiently in cost-conscious settings.